Slant-Bed CNC Lathe Build

[Karl_T]

Galil will run any servo that takes a 0 - 10 volt signal.

I play fast and loose with the volt ratings. No way will the insulation break down if you go 20 volt higher. They just don't want the extreme RPM the motor would spin up to at the higher voltage. That never happens under servo control anyway. This may shock some of the more conservative in the crowd.

why this is late:
https://www.hobby-machinist.com/threads/message-awaiting-moderator-approval.72040/#post-604425

 

[barnbwt]

Galil will run any servo that takes a 0 - 10 volt signal.

I play fast and loose with the volt ratings. No way will the insulation break down if you go 20 volt higher. They just don't want the extreme RPM the motor would spin up to at the higher voltage. That never happens under servo control anyway. This may shock some of the more conservative in the crowd.

why this is late:
https://www.hobby-machinist.com/threads/message-awaiting-moderator-approval.72040/#post-604425

Okay, increasing the voltage (within reason) makes sense; but how would one drive a 48V servo at 60V without creating an overcurrent condition --something the motors most definitely cannot tolerate? As I see it, I could get a 48V servo which runs at 2.5A, and drive it at 60V to draw ~3.1A. This puts me closer to my 200W target zone of 60V @ 3.3A, but the 25% extra current seems like it could be an issue for heat generation.

Or is the effective (not instantaneous) current limited by PWM duty-cycle, so the same 2.5A are drawn despite the 48V peak voltage during the shorter pulses?

 

[Karl_T]

Yep, you're not within specs. don't expect any warranty. I'm just sayin' I've done it.

Keep in mind I don't push my machine to max speeds and feeds so the duty cycle exceeding specs is minimal.

 

[barnbwt]

Okay, I think I found a possible brushed, DC, motor/encoder pair (it seems like these aren't wedded/integrated as often in DC servo applications for whatever reason, the way BLDC motors are; I'm guessing since the BLDC needs the encoders for its commutation, whereas the brushes do all that mechanically?)

Reliance Electro-Craft E243 Series (this example is broken, but they are a common motor on ebay)
It appears these are legacy products that have gone through several name changes (E or G, 240 or 540). But the specs I see from several different old catalogs appear the same (same peak/cont torque, same current draw, RPMs were derated to 5000krpm from 6000krpm at some point)

60VDC Max operating voltage --Check
4A, 3.1A, 2.3A cont. current --Check (this is for the A, B, and C winding, I think. The linked motor appears to be the 'A' winding; I'm limited to 3.3A)
5000-6000RPM max speed --6-7:1 ratio
55oz-in cont. torque --330-385oz-in, this seems possible a little low; my rapids may need to drop some more
Many of these are fitted with an HP HEDS-5000 type encoder (not much info on the various options, however it does appear at least some are 1000ppr, likely depending on vintage)
Many of these look to be in the sub-100$ range which is also attractive; not that different from the steppers in that they plug directly to my board amps

I *think* those parameters all fall within the specs of my on-board servo amplifiers, if I am understanding them correctly

 

[barnbwt]

Yep, you're not within specs. don't expect any warranty. I'm just sayin' I've done it.

Keep in mind I don't push my machine to max speeds and feeds so the duty cycle exceeding specs is minimal.

Yeah, I have been somewhat counting on the fact this machine is a lathe (vs. mill) helping me out in that department. Not very often you'd be rapiding fully across a 6"x6" space, both axes at the same time, against a load. Usually you'd have one axis doing all the work, the other moving slowly or holding position. And seeing how the cutting is 'automated,' I can afford to do things more slowly so long as they are consistent.

Honestly, the only time anything should see its limits pushed to an extent, would be the spindle turning as fast as it can when cutting aluminum (since it has such high surface speeds). Even then a shallow cut will be the most common action.

 

[Karl_T]

I use the E690 series of that exact motor on my mill.

It is rated 60v max. I use a 2:1 step down transformer through a bridge rectifier and then an electrolytic capacitor for a very cheap effective power supply that gives about 70 volts. I got the transformer and capacitor free, paid about 50 cents for the rectifier. Just google for dc power supply and plans for these pop right up. If you care a power transistor can easily be added to limit the voltage. I *WAS* going to do this but did not have one the weekend all the parts went together. Ran it without and nothing got even close to hot so I've never got around to adding this..

There are AMC brand servo amps on eBay for these servos that go for peanuts.

This combination works great with galil.

 

[Karl_T]

just for grins I surfed for an AMC servo amp. here's a new one:
https://www.ebay.com/itm/Advanced-M...ols")+Servo+&_from=R40&rt=nc&LH_TitleDesc=0|0

does up to 12 amps and 80 volts. Being the extremely frugal type (cheap a$$ in other words) , I watched for these and never paid $20 each

 

[barnbwt]

Well, we'll see if these Electrocraft E243 motors work out. Going by torque charts at 4000rpm, it looks like I'll have around 700lbs holding force on Z, 950lbs on X, at about an 8:1 gear ratio for both (still within belt range so I can avoid gearboxes, though they do exist). I also found a nice 60V switching power supply good up to 8A that should be overkill in practice.

Dialing back my rapid speed 'requirement' from five to seven seconds traverse on Z makes things much more reasonable as far as the 200W-class motors are concerned. Direct-drive is nice in theory because of the simplicity, but now I think I can use a belt connection to tuck the X-axis servo beneath the apron for a very clean and protected result, so that's a net improvement.

I'm still debating whether to go with a quick-change toolpost arrangment with one tool block at the lower/left corner of the top slide, or a gang arrangement. The former gives better modularity but is more labor intensive without a tool changer (one day...) the latter is simpler to implement but slightly more work to program for. Both put the center of the slide at a slightly different spot relative to the spindle axis. Any slant-bed users care to weigh in on which they prefer? Since I work around Swiss lathes all day, I keep getting weird ideas for how to arrange the tool mounts, but simplicity is probably next to godliness for this project.

 

[JimDawson]

My preference is a turret. But that would be a lot of work to implement in this case and add a lot of moving mass to the system. The next choice would be a gang tool setup. Not really difficult to program, just assign each station a tool number and and let the software take care of the offsets. In the software, the X & Z positions are set to some position, normally referenced to the X and Z home as machine coordinates. In my software both the machine and part coordinate offsets is set with a mouse click. Then that offset is applied to each tool when called. X is always relative to the spindle C/L where the tool point on C/L = X 0.0000

BTW, I don't like switching power supplies for motor power. If you don't want to build one, then something like these would be my choice. https://www.automationtechnologiesinc.com/products-page/torroidal-power-supplies/

 

[Karl_T]

I'd go gang IF you can install a long X axis. Simple and really fast to switch tools.

OTOH, I have a turret with a change plate. As long as the plate is assembled, can go back to that part in just a few minutes. Big time saver.


BTW, I've seen reports of issues with switching power supplies. No first hand knowledge.